(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



CFDP1 regulates the stability of pericentric heterochromatin thereby affecting RAN GTPase activity and mitotic spindle formation [1]

['Gokul Gopinathan', 'School Of Medicine', 'Dentistry', 'University Of Rochester', 'Rochester', 'New York', 'United States Of America', 'Qian Xu', 'Xianghong Luan', 'Thomas G. H. Diekwisch']

Date: 2024-04

The densely packed centromeric heterochromatin at minor and major satellites is comprised of H3K9me2/3 histones, the heterochromatin protein HP1α, and histone variants. In the present study, we sought to determine the mechanisms by which condensed heterochromatin at major and minor satellites stabilized by the chromatin factor CFDP1 affects the activity of the small GTPase Ran as a requirement for spindle formation. CFDP1 colocalized with heterochromatin at major and minor satellites and was essential for the structural stability of centromeric heterochromatin. Loss of CENPA, HP1α, and H2A.Z heterochromatin components resulted in decreased binding of the spindle nucleation facilitator RCC1 to minor and major satellite repeats. Decreased RanGTP levels as a result of diminished RCC1 binding interfered with chromatin-mediated microtubule nucleation at the onset of mitotic spindle formation. Rescuing chromatin H2A.Z levels in cells and mice lacking CFDP1 through knock-down of the histone chaperone ANP32E not only partially restored RCC1-dependent RanGTP levels but also alleviated CFDP1-knockout-related craniofacial defects and increased microtubule nucleation in CFDP1/ANP32E co-silenced cells. Together, these studies provide evidence for a direct link between condensed heterochromatin at major and minor satellites and microtubule nucleation through the chromatin protein CFDP1.

Funding: This study has been supported by NIH grants DE13095 and DE026198 to TGHD, and DE027686 and DE019463 to XL. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Our interest in the heterochromatin/RanGTP interface is based on a unique protein, CFDP1 (craniofacial development protein 1), originally cloned and characterized in our laboratory as an essential protein involved in cell proliferation, survival of mouse fibroblasts, and craniofacial development [ 30 , 31 ]. Other laboratories have indirectly linked CFDP1 to H2A.Z exchange based on studies in yeast [ 32 , 33 ]. Suggestive of a role in the maintenance of higher order chromatin organization, both CFDP1 and its Drosophila homologue Yeti were found to interact with HP1α [ 34 ]. The present study was designed to investigate the role of CFDP1 at the heterochromatin/RCC1 interface as it relates to the major heterochromatin components HP1α and H2A.Z and its potential implication on RCC1-mediated Ran activity. Our data suggest that CFDP1 plays a crucial role in the regulation of RanGTP levels and related chromosomal microtubule nucleation through specific modulation of heterochromatin components at the centric and pericentric chromatin domains.

Early developmental studies using mutant mice have demonstrated that the variant histone H2A.Z is necessary for proper chromosome segregation and for the retention of the major heterochromatin protein HP1α at the PCH [ 23 – 25 ]. H2A.Z is present at both centric and PCH and contributes to centromere identity by maintaining optimal CENP-A levels at the centromere of mitotic chromosomes [ 25 , 26 ]. At centric heterochromatin, H2A.Z nucleosomes paired with dimethylated H3K4 are interspersed between CENP-A subdomains, while at the PCH, H2A.Z nucleosomes are paired with trimethylated H3K4 histones [ 27 ]. Linking H2A.Z to the major heterochromatin protein HP1α, H2A.Z has been demonstrated to induce higher order chromatin fiber folding mediated by HP1α, while also interfacing with H3K9me3 in regulating HP1α binding to nucleosomes [ 24 , 27 ]. HP1α homodimers directly bind to H3K9me3 heterochromatin marks and to Suv39h1/2, the histone methyltransferase that deposits H3K9me3, thereby contributing to both the maintenance of heterochromatin structure and the establishment of histone methylation patterns [ 28 , 29 ].

Further underscoring the involvement of heterochromatin in the facilitation of cell division, centric and pericentric regions of centromeric heterochromatin through minor and major satellite tandem repeat regions play distinct roles during spindle attachment and chromatin cohesion [ 16 , 17 ]. The minor satellites are comprised of approximately 600 kb of 120-bp AT-rich monomers and coincide with the centric region of mouse chromosomes, whereas the major satellites are located pericentrically to the minor satellites and are composed of 6 megabases of 234-bp monomers [ 18 – 20 ]. In interphase nuclei, major satellites from different chromosomes cluster together to form chromocenters with the corresponding minor satellites located as separate entities in the periphery [ 14 , 21 ]. Distinguishing minor and major satellites, CENP-A is associated only with minor satellites, whereas HP1α specifically accumulates on the major satellites [ 14 ]. This difference is particularly critical during mitosis, when the CENP-A rich centric domain serves as the site for kinetochore formation and spindle microtubule attachment, while the HP1α containing pericentric domain is implicated in sister chromatid cohesion [ 22 ].

Chromatin is a highly organized complex of DNA and proteins consisting of 2 distinct structural domains, the open and active chromatin regions characterized as euchromatin and the highly condensed, gene-poor, and less active chromatin regions called heterochromatin [ 8 ]. Supportive of a distinct role of heterochromatin in mitosis, the highly condensed heterochromatin is known to participate in sister chromatid cohesion and chromosome segregation, likely due to its structural properties ( Fig 1 ) [ 9 ]. Other pieces of evidence that link heterochromatin to individual aspects of mitosis include the recruitment of the cohesin protein complex for sister chromatid cohesion [ 10 , 11 ] and its scaffolding function during kinetochore assembly [ 12 , 13 ]. In most eukaryotes including mammals, centromere identity and function are marked by epigenetic components such as hypoacetylated histones, histone H3 trimethylated at lysine 9 (H3K9me3), and histone variants [ 14 , 15 ]. The relative distribution of these chromatin components within the centromeric heterochromatin further specifies 2 adjacent subdomains featuring distinct functional properties: the centric and pericentric heterochromatin (PCH) domains [ 14 ]. The centric chromatin is marked by the histone H3 variant CENP-A, whereas the surrounding pericentric domain is enriched for several H2A variants including H2A.Z, while it is completely devoid of CENP-A ( Fig 1 ) [ 14 ].

Mitotic cell division through spindle assembly and chromosome segregation is essential to eukaryotic life. Yet, the mechanisms by which the mitotic spindle achieves this extraordinary feat of segregating pairs of chromatids into 2 daughter cells are poorly understood [ 1 ]. Classic studies have linked bipolar spindle assembly to a “search and capture” process, by which duplicating centrosomes nucleate microtubules which then capture kinetochores and pull opposing spindle ends apart [ 2 ]. This original working hypothesis has been amended to include the celebrated chromatin-mediated spindle assembly model, which suggests that bipolar spindles assemble independent from centrosomes and kinetochores but rather in response to mitotic chromatin and its effect on the local environment to promote microtubule nucleation and stabilization [ 3 , 4 ]. Subsequent studies have identified RanGTP gradients generated by the chromatin-associated factor RCC1 as major local triggers of spindle assembly, linking chromatin with microtubule nucleation [ 5 – 7 ]. Yet, how does chromatin facilitate the formation of these all-important RanGTP gradients at the very onset of mitotic spindle formation?

2. Results

2.1. CFDP1 regulates pericentric heterochromatin compaction by modulating HP1α and H3K9me3 levels in a reversible fashion Previous studies in Drosophila have colocalized the CFDP1 homologue Yeti with heterochromatin loci on polytene chromosomes and reported interactions between the heterochromatin marker protein HP1α and CFDP1 [34,35]. To determine whether CFDP1 plays a structural role in heterochromatin organization in mammals, we first examined its subcellular localization in NIH3T3 fibroblasts using immunofluorescence assays. Our experiments revealed a punctate pattern of CFDP1 distribution within the cell nucleus where CFDP1 foci colocalized with DAPI foci and also stained positive for the core heterochromatin protein HP1α (Fig 2A) and the repressive histone modification H3K9me3 (Fig 2B). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Analysis of chromatin compaction mediated by CFDP1. (A, B) Immunofluorescence analysis of CFDP1 colocalization with the heterochromatin protein HP1α and H3K9me3 histone modification. (C, D) Immunofluorescence analysis for CFDP1 association with the PCH in TSA treated NIH3T3 cells. (E, F) Representative DNA gel electrophoresis image comparing MNase digestion profile for chromatin isolated from (E) NIH3T3 cells treated with control siRNA (con si) and CFDP1 siRNA (CFDP1 si) and in (F) NIH3T3 cells overexpressing full-length CFDP1 (pSFCMV-CFDP1) or a control vector (pSFCMV). Chromatin was digested with MNase (at indicated concentrations) and the isolated DNA was run on 1.5% agarose gels and stained with ethidium bromide. DNA marker standards were run alongside for reference. (G, H) Electropherogram comparison for DNA staining intensity signal (gray value) versus fragment size distribution obtained from DNA electrophoresis of MNase digested chromatin in (E) con/CFDP1 siRNA-treated cells and in (F) vector/CFDP1 overexpressing cells. Peaks corresponding to mononucleosomes (mono) and dinucleosomes (di) are demarcated horizontally. Gray value plots were obtained from Image J analysis of a representative DNA run (n = 3). (I, J) Heterochromatin structure is responsive to CFDP1 levels. Chromatin levels of core heterochromatin proteins, HP1α, H3K9me3, and H2A.Z were analyzed in chromatin extracts prepared from (I) CFDP1 siRNA-treated cells and (J) cells overexpressing CFDP1. (K, L) Normalized fluorescence intensity obtained from FRAP assay for HP1α at PCH foci in (K) CFDP1 siRNA-treated cells or (L) CFDP1 overexpressing cells. EGFP-tagged HP1α expressing NIH3T3 cells were subjected to CFDP1 siRNA or overexpression and processed for FRAP analysis. Raw data was normalized using a double-normalization method and the mean normalized curve is plotted (S1 Data). (M) Quantitation of FRAP metrics. Calculation of half life (t1/2) and mobile fraction (%) for HP1α at PCH in cells treated with CFDP1 siRNA and in CFDP1 overexpressing cells. Molecular weights of detected proteins are indicated next to the immunoblots. s, seconds; U, MNase units; bp, base pairs; kDa, kilo Dalton. *** p = 0.001. Scale bar = 5 μm. CFDP1, craniofacial development protein 1; FRAP, fluorescence recovery after photobleaching; PCH, pericentric heterochromatin; TSA, trichostatin A. https://doi.org/10.1371/journal.pbio.3002574.g002 Cells were treated with TSA (trichostatin A), resulting in fragmentation and dissolution of higher order heterochromatin structure as visualized by the appearance of a large number of smaller-sized DAPI foci (Fig 2C). Immunofluorescence assays in TSA treated cells identified a large number of smaller-sized CFDP1 foci that colocalized with the fragmented heterochromatin, suggesting that CFDP1 is a stable and core component of heterochromatin (Fig 2C). To examine the physiological effects of CFDP1 on chromatin, we performed chromatin accessibility assays on NIH3T3 cells treated with siRNA against CFDP1 and on NIH3T3 cells stably expressing 3XFLAG tagged CFDP1. Chromatin isolated from treated cells was digested with increasing concentrations of micrococcal nuclease (MNase) to reveal its nucleosomal organization. These assays identified higher chromatin accessibility in CFDP1 knockdown cells compared to control siRNA-treated cells (Fig 2E), suggesting that CFDP1 plays a key role in the structural organization of chromatin. This difference was especially prominent at the highest concentration of MNase used (2.0 U), where most of chromatin was digested to the length of mono- or di-nucleosomes in CFDP1 knockdown cells compared to control cells, which displayed higher-level nucleosomal configurations (Fig 2E). Conversely, overexpression of CFDP1 in NIH3T3 cells led to the opposite effect, with a decrease in MNase chromatin accessibility (Fig 2F), indicating CFDP1 induces chromatin compaction. Together, these studies suggest that CFDP1 is involved in the regulation of chromatin accessibility and structure. We hypothesized that such large-scale changes in chromatin accessibility are due to a direct modulation of heterochromatin structure by CFDP1, since heterochromatin plays a key role in higher order chromatin structural organization, and loss of heterochromatin proteins impacts chromatin accessibility [36]. In support of our hypothesis, we determined that crude chromatin extracts from CFDP1 siRNA-treated cells (siRNA treatment resulted in 60% reduction in CFDP1 protein levels compared to controls) demonstrated significantly lower levels of the core heterochromatin protein HP1α and the associated repressive histone modification H3K9me3 compared to control siRNA-treated cells (Fig 2I). Moreover, chromatin levels of the histone variant H2A.Z were also reduced upon CFDP1 knockdown. Conversely, chromatin from CFDP1 overexpressing cells (2.1-fold increase in CFDP1 protein levels) displayed higher levels of HP1α, H3K9me3, and H2A.Z compared to a vector control (Fig 2J). These results demonstrated a direct and dose-dependent role for CFDP1 in the maintenance of a stable heterochromatin structure. We then set out to determine whether the structural changes in heterochromatin upon CFDP1 knockdown or overexpression are due to differences in the mobility of core heterochromatin proteins. We performed fluorescence recovery after photobleaching (FRAP) analysis of GFP tagged HP1α molecules specifically at the PCH, a domain which is highly enriched for HP1α [37], and hence might be affected by CFDP1 levels. Our assays demonstrated a higher mobility and faster recovery kinetics for HP1α at the PCH in CFDP1 knockdown cells compared to control siRNA-treated cells (Figs 2K, S1A and S1B) supporting our previous results of increased chromatin accessibility upon CFDP1 depletion. siRNA mediated depletion of CFDP1 resulted in a significant reduction of HP1α-GFP recovery half-life (t1/2) and a small increase in the mobile fraction (Fig 2M). Consistent with previous reports [36], and indicative of HP1α as a highly mobile chromatin protein, our assays also demonstrated a high turnover of HP1α at PCH with fluorescence recovery on the seconds scale. In addition, we confirmed our finding of increased chromatin accessibility in CFDP1 knockdown cells by performing FRAP assay in Histone H1f1-GFP expressing cells (S1C Fig, S1D Fig and S1 Data). Interestingly, HP1α dynamics at the PCH was slightly decreased upon CFDP1 overexpression (pSFCMV CFDP1) compared to control vector (pSFCMV) expression (Fig 2L and 2M), further validating the dual role of CFDP1 on the PCH affecting both heterochromatin protein levels and mobility.

2.2. CFDP1 is enriched at DNA repeat elements in the mouse genome and is preferentially associated with H2A.Z nucleosomes in vitro Based on our findings demonstrating a functional role for CFDP1 in the structural organization of chromatin especially at the PCH, we asked whether CFDP1 executes this role by directly binding to distinct chromosomal regions. To identify CFDP1 binding sites in the mouse genome, we performed high-throughput sequencing of CFDP1 associated chromatin fractions isolated from FLAG immunoprecipitations of chromatin from 3XFLAG-CFDP1 expressing NIH3T3 cells. Based on the direct effect of CFDP1 on histone variant H2A.Z chromatin levels (Fig 2I and 2J) along with the established role of the CFDP1 yeast homologue Swc5 in histone variant exchange [33,38], we compared CFDP1 bound chromatin regions with genomic regions enriched for H2A.Z. This analysis identified a very limited number of peaks (6,430 peaks identified by Homer) for CFDP1 binding, indicating that CFDP1 is not ubiquitously bound to chromatin but rather that significant CFDP1 levels are present only at a few discrete genomic regions within the mouse genome. As expected, enrichment for H2A.Z was observed at a large number of genomic sites (52,234 peaks identified by Homer). Our analysis identified a discrete overlap between CFDP1 and H2A.Z only at a few peaks (413 peaks in common, S3 Table). Genome ontology (GO) of CFDP1 enriched peaks identified a large proportion of DNA repeat regions indicating that CFDP1 preferentially binds to repeat elements in the mouse genome, including simple repeats, LINEs, SINEs, LTRs, and satellite repeats (S1 Table). Interestingly, GO analysis for H2A.Z enriched peaks also identified a large number of DNA repeat elements in addition to the known enrichment of H2A.Z at promoter regions (S2 Table). Remarkably, CFDP1 and H2A.Z peaks colocalized at discrete genomic regions which mapped to DNA repeat elements including simple repeats, LINE, LTR, SINE, and satellites (Figs 3A and S2A). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Characterization of CFDP1-H2A.Z interaction. (A) ChIP-seq analysis for CFDP1 binding in the mouse genome. IGV browser snapshot visualizing genomic occupancy of 3X FLAG tagged CFDP1 and H2A.Z at DNA repeat elements in the mouse genome. CFDP1 and H2A.Z display similar enrichment profiles at several repeat elements in different chromosomal regions as indicated. CFDP1 ChIP-seq was performed following FLAG immunoprecipitation of NIH3T3 cells stably expressing full-length 3XFLAG-CFDP1. (B) IGV browser snapshot for H2A.Z enrichment on DNA repeat elements at indicated chromosomal regions in control and CFDP1 siRNA-treated NIH3T3 cells. Peaks demonstrating pronounced decrease in H2A.Z enrichment after CFDP1 siRNA treatment are marked by an asterisk (*). (C) siRNA-mediated knockdown of CFDP1 (CFDP1 si) and SRCAP (SRCAP si) decreased chromatin levels of H2A.Z. (D) siRNA-mediated knockdown of CFDP1 expression does not affect total levels of H2A.Z protein. (E) CFDP1 depletion led to a remarkable reduction in chromatin incorporation of exogenously supplied FLAG-H2A.Z compared to control siRNA (con si)-treated cells (<2.5-fold). (F) CFDP1 preferentially interacts with H2A.Z nucleosomes in vitro. Full-length CFDP1 protein was added to H2A canonical nucleosomes or to H2A.Z nucleosomes bound to streptavidin magnetic beads. Nucleosome bound fraction was purified and probed for CFDP1 and histones as indicated. (G) Canonical H2A or H2A.Z recombinant nucleosomes were incubated with HIS magnetic beads pre-bound to HIS-CFDP1 protein. The bead bound fraction was purified and probed for CFDP1 and histone H4 proteins. (H) The acidic N terminus of CFDP1 specifically interacts with H2A/H2B dimers. HIS tagged full-length CFDP1 or fragments of CFDP1 were incubated with histone dimers and subjected to immunoprecipitation using HIS magnetic beads. Immunoprecipitates were probed for CFDP1 and histone proteins as indicated. Molecular weights of detected proteins are indicated next to the immunoblots. CFDP1, craniofacial development protein 1; ChIP, chromatin immunoprecipitation; IP, immunoprecipitation; kDa, Kilo Dalton; nuc, nucleosome. https://doi.org/10.1371/journal.pbio.3002574.g003 Based on the decrease in H2A.Z chromatin levels upon CFDP1 knockdown, we next compared H2A.Z enrichment levels at various DNA repeats in control and CFDP1 siRNA-treated cells by sequencing H2A.Z ChIP enriched DNA fragments. This analysis revealed a remarkable reduction in the levels of H2A.Z enrichment at several DNA repeat elements in CFDP1 siRNA-treated cells compared to control siRNA-treated cells (Fig 3B and S4 Table). Furthermore, GO analysis for these H2A.Z enriched peaks from control siRNA-treated cells identified repeat elements as a prominent annotation term (S5 Table). Two experimental pieces of evidence prompted us to further investigate whether there was a molecular interaction between CFDP1 and H2A.Z: (i) a colocalization between CFDP1 and H2A.Z enriched peaks at genomic repeats based on our ChIP-seq studies; and (ii) the decrease in H2A.Z chromatin levels upon CFDP1 knockdown. Similar to the role played by its yeast homologue swc5, we demonstrate that CFDP1 is essential for the incorporation of H2A.Z into chromatin (Fig 3C, 20% reduction). As a comparison, knockdown of SRCAP, the catalytic component of SRCAP complex necessary for H2A.Z incorporation in nucleosomes led to extensive loss of H2A.Z from chromatin (Fig 3C). We interpret the decrease in H2A.Z chromatin levels upon CFDP1 knockdown (CFDP1 proteins levels decreased by 55%) as a result of defective histone variant incorporation because CFDP1 siRNA treatment did not affect total H2A.Z protein levels (Fig 3D). Furthermore, there was a substantial decrease in FLAG-H2A.Z (exogenously transfected H2A.Z) incorporation into the chromatin of CFDP1 depleted cells (2.5-fold less) compared to control cells as revealed by immunoblot assays using an anti-FLAG antibody (Fig 3E). Together, these data demonstrate a close molecular association between CFDP1 and H2A.Z, wherein CFDP1 is enriched at a small number of genomic repeats along with H2A.Z, while CFDP1 also plays an important role in H2A.Z incorporation. Further confirming the molecular link between CFDP1 and H2A.Z incorporation, in vitro binding assays revealed that full-length CFDP1 has a 3.3-fold (normalized to H4 levels) higher affinity for H2A.Z variant containing nucleosomes over H2A containing canonical nucleosomes (Fig 3F). This finding was verified by HIS tag immunoprecipitation of HIS-CFDP1 protein incubated with H2A or H2A.Z nucleosomes demonstrating that CFDP1 was almost exclusively bound to H2A.Z nucleosomes (Fig 3G). We then compared individual CFDP1 domains, the N-terminus (aa 1–150), center fragment (aa 99–199), C-terminus (aa 150–295), and the BCNT domain (aa 218–295) to ask which domain of the CFDP1 molecule interacts with individual histones. These studies demonstrated that among the CFDP1 fragments, only the acidic N-terminus interacted specifically and strongly with H2A/H2B dimers (Fig 3H). On the other hand, all CFDP1 fragments interacted with H2A.Z/H2B dimers and the H3/H4 tetramers in independent assays (Fig 3H).

2.3. CFDP1 regulates heterochromatin state by stabilizing HP1α and H3K9me3 at major satellites and CENPA at minor satellites Our biochemical assays linking CFDP1 to chromatin occupancy of HP1α and H2A.Z suggested a direct involvement of CFDP1 in regulating PCH structure since HP1α and H2A.Z along with H3K9me3 are essential for PCH structural stability [39]. We therefore conducted a detailed molecular and spatial profiling of CFDP1 enrichment among minor and major satellites within mouse heterochromatin. In mouse nuclei, PCH from different chromosomes comprising major satellite DNA repeat elements coalesces to form chromocenters, with the corresponding minor satellites localized to the peripheral margins [14]. Chromocenters are visualized as DAPI dense foci in mouse nuclei and our immunofluorescence localization assays demonstrated CFDP1 enrichment at these foci in most of the nuclei. CFDP1 staining was predominantly observed along peripheral margin of the chromocenter (Fig 4A). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Pericentric heterochromatin structural integrity is disrupted in the absence of CFDP1 due to loss of core heterochromatin protein, HP1α and H3K9me3 marks. (A) Immunofluorescence analysis illustrates a punctate CFDP1 localization pattern along the peripheral margins of chromocenters. (B, C) Immunofluorescence coupled with FISH assay for CFDP1 colocalization at the (B) major and (C) minor satellite repeats. CFDP1 was localized using HIS antibody in cells expressing HIS-CFDP1 and then subjected to in situ hybridization with minor and major satellite probes. (D) ChIP assay for H2A.Z occupancy at DNA repeat elements in chromatin from control, CFDP1 and SRCAP siRNA-treated NIH3T3 cells as indicated. H2A.Z levels were decreased at all repeat elements tested. (E) CFDP1 knockdown leads to decreased incorporation of exogenously transfected FLAG-H2A.Z at DNA repeat elements. ChIP assay was performed in cells following simultaneous knockdown of CFDP1 and transfection of FLAG-H2A.Z. (F) Decreased H3K9me3 enrichment at major satellite repeats upon CFDP1 siRNA treatment. (G) ChIP assay demonstrating significant reduction in HP1α enrichment at major satellite repeats after CFDP1 knockdown. (H) CFDP1 knockdown by siRNA resulted in a significant reduction of CENP A binding to the minor satellite repeats. ChIP PCR, n = 4 from 3 independent ChIP experiments (error bars = ± SEM, S2 Data). min SAT, minor satellites; maj SAT, major satellites; n.s, not significant. p value (* < 0.05, ** < 0.01, *** < 0.001). CFDP1, craniofacial development protein 1; ChIP, chromatin immunoprecipitation; FISH, fluorescent in situ hybridization. https://doi.org/10.1371/journal.pbio.3002574.g004 This distinctive localization pattern of CFDP1 within the PCH suggested that CFDP1 might be localized at the major and minor satellite repeats. We therefore performed fluorescent in situ hybridization (FISH) assays coupled with immunofluorescence to precisely determine CFDP1 localization with respect to minor and major satellite repeat elements. These assays confirmed the close association between CFDP1 and satellite repeat elements within mouse heterochromatin (Fig 4B and 4C). CFDP1 was predominantly observed along the outer margins of the larger major satellite foci (Fig 4B) and also colocalized with several smaller foci corresponding to minor satellite regions (Fig 4C) in our FISH colocalization assays. As with our immunofluorescence studies, CFDP1 staining was visible within the core region of major satellite stained PCH foci, although these levels were less intense when compared to CFDP1 staining at the periphery of major satellite foci (Fig 4B). The essential role of CFDP1 related to H2A.Z incorporation, its regulation of HP1α and H3K9me3 together with its distinctive localization at PCH prompted us to hypothesize that altered chromatin accessibility upon CFDP1 modulation might be directly related to local chromatin changes at minor and major satellite repeats. To verify whether changes in CFDP1 levels affect chromatin at minor and major satellites, we performed ChIP studies to access chromatin levels of H2A.Z at minor and major satellite repeat elements in control and CFDP1 knockdown cells. Based on our ChIP-seq analysis, we tested H2A.Z levels at other repeat elements including, TLC, LINE1 ORF1, and SINEs. Corroborating the decrease in H2A.Z chromatin levels upon CFDP1 knockdown, our analysis demonstrated a significant decrease in H2A.Z occupancy (using both endogenous and exogenously supplied H2A.Z) at all DNA repeat elements tested, including the minor and major satellite repeats in CFDP1 siRNA-treated cells (Fig 4D and 4E). As expected, depletion of the CFDP1-associated SRCAP complex resulted in a highly dramatic decrease in H2A.Z enrichment at all repeat elements (Fig 4D and 4E). Interestingly, this decrease in H2A.Z enrichment upon CFDP1 knockdown was associated with a small but significant increase in the expression levels of minor and major satellite repeats (S2B Fig). We next tested whether levels of the PCH components, HP1α and H3K9me3 were affected at major satellite repeats upon CFDP1 depletion. H3K9me3 along with HP1α are known to be preferentially enriched at the PCH, where they perform an essential function in the maintenance of structural stability [37]. Our analysis demonstrated a substantial decrease in H3K9me3 enrichment at the major satellite repeats upon CFDP1 depletion, while enrichment at the minor satellite repeat elements was not affected (Fig 4F). The overall decreased enrichment for H3K9me3 at minor satellite repeats and the absence of any difference in enrichment upon CFDP1 depletion was expected because H3K9me3 is preferentially present at major satellite repeats compared to minor satellites [27]. Adding to loss of H3K9me3, ChIP analysis also identified a small but significant decrease in chromatin levels of HP1α at major satellite repeats (Fig 4G). CFDP1 depletion not only affected the major pericentric heterochromatin components, HP1α and H3K9me3, but also caused a significant loss of the CENP-A nucleosome variant at minor satellite repeats, pointing toward a major role for CFDP1 at the centric heterochromatin (Fig 4H). Together, these data further solidify the role of CFDP1 in the maintenance of heterochromatin structure and stability.

2.4. CFDP1 is specifically targeted to the pericentric heterochromatin domain at mid-late stage S phase and is essential for its timely replication Our experiments so far have demonstrated a critical role for CFDP1 in PCH assembly. We thus decided to test whether CFDP1 knockdown affects PCH duplication since pericentric heterochromatin structure and cell-cycle progression are closely related. Defects in cell cycle progression have been previously documented in cells lacking CFDP1 [30,31,40]. PCH usually replicates during mid-late S phase in specific units termed pericentric heterochromatin duplication bodies (pHDBs) [41]. pHDBs are visualized upon DAPI staining as characteristic ring-like structures surrounding the periphery of heterochromatin [41]. Here, we performed immunofluorescence localization studies on cells expressing HIS tagged CFDP1, and cells were scored for different phases of the cell cycle using a dual labeling strategy involving AURORA B and EdU [42]. Interphase cells were categorized as either G1 phase (EdU and AURORA B negative), S phase (EdU positive), or G2 phase (EdU negative, AURORA B positive) based on immunofluorescence staining for AURORA B and the Click-iT reaction that reveals EdU incorporation in replicating DNA. S phase cells were further designated as early S, Mid-late S, and very late S phase based on EdU distribution patterns. Our analysis revealed that CFDP1 foci colocalized with DAPI foci throughout interphase, suggesting that CFDP1 is preferentially localized as part of heterochromatin throughout the interphase (Fig 5A). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. CFDP1 is associated with replicating pericentric heterochromatin during mid-late S-phase and CFDP1 depletion results in delayed pericentric heterochromatin duplication. (A–F) Nuclear localization of CFDP1, Aurora B and EdU label in interphase stage 3T3 cells. Stages within interphase were resolved based on immunofluorescent staining patterns for Aurora B and EdU labeling as: G1 (negative for AURORA B and EdU), S (positive for EdU), and G2 (positive for Aurora B and negative for EdU). S phase cells were further scored as early, mid-late, and very late stages based on EdU distribution patterns. Throughout interphase, CFDP1 is colocalized with DAPI dense foci (heterochromatin). CFDP1 foci were excluded from replicating DNA in early S (B) and very-late S phase (E) nuclei while it was closely associated with replicating pericentric heterochromatin at the mid-late S phase (C, D). (G–I) Delayed S phase progression in CFDP1 siRNA-treated cells. (G) Quantitative analysis of EdU labeled cells in early, mid-late, or very late stages of S phase in asynchronously dividing, control and CFDP1 siRNA-treated cells. (H, I) Quantitative analysis for percentage of mid-late stage nuclei among all S phase cells in control and CFDP1 siRNA-treated synchronized cells released from a double thymidine block. For this study, cells were labeled with EdU after 4 h and 6 h of release from S phase block. Note the higher percentage of mid-late stage, S phase nuclei in CFDP1 siRNA-treated cells in G, H, and I. Percentage of S phase nuclei represent the mean of 3–5 independent experiments (n > 200 S phase nuclei per experiment, error bars = ± SEM, S3 Data). Thy, Thymidine. p value *** < 0.001. Scale bar = 5 μm. CFDP1, craniofacial development protein 1. https://doi.org/10.1371/journal.pbio.3002574.g005 Cells in G1, Early S, very late S, and G2 phases displayed intense staining for CFDP1 at pericentric heterochromatin duplication bodies, while CFDP1 foci at mid-late S stage were smaller and rather distributed within the nucleus. Furthermore, CFDP1 foci were clearly excluded from replicating DNA in early S and very late-stage S phase cells, while they were closely associated with the replicating DNA in mid-late S phase cells (Fig 5A to 5F). The presence of CFDP1 in mid-late stage S-phase heterochromatin suggests that CFDP1 functions as part of the structural rearrangements during pericentric heterochromatin replication that occur during the cell cycle transition [43]. The tight association between CFDP1 and actively replicating PCH DNA during mid-late S phase further supports a structural role for CFDP1 at the PCH domain, suggesting that CFDP1 plays a role in the re-assembly of higher order chromatin structure at the PCH immediately after DNA replication. Based on the close association between CFDP1 and S-phase heterochromatin, we tested whether CFDP1 was essential for a timely transition through the S phase in NIH3T3 fibroblast cells. S phase nuclei from control and CFDP1 siRNA-treated cells were scored based on EdU incorporation and AURORA B staining. Our experiments in asynchronously dividing siRNA-treated NIH3T3 fibroblasts revealed a delay in progression through the S subphases with a significantly higher percentage of cells in the mid-to-late phase of CFDP1 depleted cells (Fig 5G). These findings were further confirmed in synchronously dividing cells following release from thymidine arrest with a significantly higher proportion of cells at the mid-late phase in CFDP1 siRNA-treated cells compared to control siRNA-treated cells (Fig 5H and 5I).

2.5. CFDP1 depleted cells revealed chromosome segregation and spindle defects Based on previous reports identifying CFDP1 as a regulator of cell division and our own results documenting S phase delays in CFDP1 depleted cells, we investigated the effects of CFDP1 knockdown during mitosis. Immunofluorescence experiments in CFDP1 siRNA-treated NIH3T3 mitotic cells exhibited a wide range of defects in the segregation of chromosomes which included high frequency of lagging chromosomes, disorganized chromosome congression at metaphase plates, and anaphase chromosome bridges (top 3 rows) compared to control siRNA (con)-treated cells (bottom row) (Fig 6A). Importantly, CFDP1 knockdown cells demonstrated striking defects in spindle organization including many multi-pole spindles (Fig 6A, multi pole), hinting to an important role for CFDP1 in microtubule organization. To rule out potential side effects due to siRNA transfection, an alternative Cfdp1 gene knockout strategy was carried out in 4-OHT (4-Hydoxytamoxifen—a metabolite of Tamoxifen) treated inducible mouse embryonic fibroblasts (MEFs) generated from Rosa26 Cre/Cfdp1-/flox embryos. This strategy was associated with a substantial decrease in CFDP1 expression after 72 h of 4-OHT incubation in a concentration-dependent and time-dependent fashion validating our CFDP1 knockout model (Fig 6B). More importantly, we were able to recapitulate the chromosome segregation defects observed in CFDP1 depleted NIH3T3 fibroblasts in Cfdp1 KO inducible MEFs upon treatment with 4-OHT (Fig 6C and 6D). The high incidence of chromosome segregation defects and spindle structural defects in CFDP1 siRNA-treated cells demonstrated a crucial role for CFDP1 in regulating microtubule dynamics and spindle stability during mitosis. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. Chromosome segregation defects and spindle defects in CFDP1 depleted cells. (A) Representative immunofluorescence experiments documenting chromosome segregation defects and spindle morphology in NIH3T3 cells treated with control siRNA (con) and CFDP1 siRNA. Mitotic defects in CFDP1 siRNA-treated cells include lagging chromosomes, multi pole spindles, and chromatin bridges (top 3 rows). (B) Immunoblot analysis demonstrating knockdown of CFDP1 protein levels in control and inducible MEFs after 72 h of 4-OHT induction at indicated concentrations (nM, top). Treatment of inducible MEFs with 250 nM 4-OHT resulted in significant decrease in CFDP1 protein levels over a 76-h time point (bottom). (C, D) Chromosome segregation defects in Cfdp1 conditional knockout MEFs. Representative immunofluorescence analysis for tubulin in uninduced MEFs (C) and 4-OHT induced MEFs (D). Mitotic substages are mentioned for MEF immunofluorescence experiments. DNA is visualized using DAPI. All images are representative of more than 3 independent experiments with 70–100 cells imaged per experimental condition. CFDP1, craniofacial development protein 1; MEF, mouse embryonic fibroblast. https://doi.org/10.1371/journal.pbio.3002574.g006

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002574

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.